Toward a predictive model of suicidal ideation and behavior: some preliminary data in college students

Within a multivariate statistical design, the relationship of various interpersonal, emotional, and cognitive variables to suicidal ideation and behavior in college students was examined. A total of 158 subjects (58 males, 100 females) completed self-report measures of life stress, faulty cognitions, loneliness, depression, hopelessness, family cohesiveness, adaptive reasons for living, and suicidal ideation and behavior. The results of a multiple-regression analysis, forward-inclusion algorithm, indicated that a predictive equation consisting of loneliness, irrational beliefs, and low adaptive reasons for living best accounted for suicidal behavior scores. To determine the basic structures and power of the predictor variables under study, a factor analysis and composite regression were employed, resulting in a predictive equation consisting of three factors: Social/Emotional Alienation, Cognitive Distortions, and Deficient Adaptive Resources. Results are discussed in terms of an interactional model of suicidal behavior.     

The psychology of self-esteem: Promise or peril?

The psychology of self-esteem bypasses consequential issues of everyday behavior and moral responsibility in favor of pursuing a global goal of enhanced self-esteem on the assumption that happiness, success, and responsible behavior will automatically follow. The goal is vague and ways of achieving it are confused and conflicting. Greater attention to concrete tasks and challenges and sensitivity to moral issues could help make therapy more clear, effective, and socially useful.     

A microcomputer-dependent statistics and design course

A one-semester course in statistics and research design is described. All computations are carried out by a Macintosh computer. No hand calculation is expected of the students, and little class time is spent on computational procedure. The course covers much more statistical content than is the norm for a one-semester course, and student understanding of the logic of the material and its application is enhanced.     

Autism and Accuracy in Adolescent Perceptions of Friends' Sexual Attitudes and Behavior

This study uses data from adolescents matched with data from their friends to examine whether reference group effects on adolescent sexuality should be thought of as operating through the sharing of norms or the modeling of behavior. We observe that perception of attitudes has such a trivial reality component and such a strong autistic component that it cannot serve as a mechanism of peer influence. Perception of sexual behavior has a reality component which is as large as its autistic component, and large enough to serve as a route of influence. We reach two conclusions: (1) Reference group effects on adolescent sexual behavior probably work through behavior modeling rather than through normative influence. (2) Studies which infer peer influence on sexual behavior using only perception of peers will arrive at erroneous conclusions.